Artificial intelligence giant OpenAI CEO Sam Altman announced that he will launch GPT-4.5 and GPT-5 models within a few months, integrating and streamlining his AI models to address the current overly complex product roadmap.
Ultraman announced this on social media X, and OpenAI realized that its model and product had become more complex. The upcoming GPT-4.5 will be the last non thought chain model and will integrate the O3 inference model into the GPT-5 model. GPT-4.5 will take several weeks to go online, while GPT-5 will take several months.
In addition, Ultraman stated that GPT-5 will be deployed to all users, even the free version can be used. Plus subscribers can use higher specification versions, while Pro subscribers can obtain even higher-level versions. The model will combine features such as voice, search, and deep research.
According to foreign media reports, OpenAI's move is seen as a response to the rise of its mainland competitor DeepSeek. The new strategy seems to no longer require users to pay $200 per month for "exclusive or advance use" of the Pro version's features, but to allow as many people as possible to use them, in order to defeat competitors with market share and user experience.
With the emergence of DeepSeek, OpenAI's position on the throne of artificial intelligence is being challenged. DeepSeek creates AI models at an extremely low price and provides them as chatbots for free, which not only shakes the global stock market but also questions the business strategies of OpenAI and other AI companies, proving that AI models can be made cheaper and easier.
Source: China Times News